Temporal and spatial distribution of pedestrians in subway evacuation under node failure by multi-hazards

The metro has become an important component of the urban traffic system. Due to the dense crowd, if sudden accidents occurred during rush hours in station, such as fires, explosions or natural disasters, the casualty and injury would be very enormous. Accidents happening in the underground station are prone to trigger domino effect, causing complex multi-hazards consequence. Therefore, more attention should be focused on the crowd evacuation strategy in underground stations. In this paper, for studying the temporal and spatial distribution characteristics of evacuation in dense subway stations under node failure caused by multi-hazard coupling conditions, a large-scale evacuation model was established by Massmotion based on the social-force, combined with the on-site measured data of Nanjing Xinjiekou subway station, one of the largest stations in Asia. A series of key node failures in the station that represents the bad impact of multi-hazards is set in the model. Through analyzing the evacuation time and the occupant density, etc, the results show that if pedestrians tend to escape by the lowest cost, it will cause extreme pressure on some narrow region. If those nodes fail unfortunately, the global evacuation efficiency will be seriously affected. Yet, an excessively scattered targets strategy obviously wastes traffic capacity and delay evacuation process. Therefore, it is necessary to balance the global evacuation time and local density in the process of crowd evacuation. The current conclusion will be useful to prompt the administrators to formulate a reasonable evacuation strategy and disperse the regional evacuation risk.